The baseline model

read image

import os
img_dir = '/tmp/nst'
if not os.path.exists(img_dir):
    os.makedirs(img_dir)
!wget --quiet -P /tmp/nst/ https://upload.wikimedia.org/wikipedia/commons/d/d7/Green_Sea_Turtle_grazing_seagrass.jpg
!wget --quiet -P /tmp/nst/ https://upload.wikimedia.org/wikipedia/commons/0/0a/The_Great_Wave_off_Kanagawa.jpg
!wget --quiet -P /tmp/nst/ https://upload.wikimedia.org/wikipedia/commons/b/b4/Vassily_Kandinsky%2C_1913_-_Composition_7.jpg
!wget --quiet -P /tmp/nst/ https://upload.wikimedia.org/wikipedia/commons/0/00/Tuebingen_Neckarfront.jpg
!wget --quiet -P /tmp/nst/ https://upload.wikimedia.org/wikipedia/commons/6/68/Pillars_of_creation_2014_HST_WFC3-UVIS_full-res_denoised.jpg
!wget --quiet -P /tmp/nst/ https://upload.wikimedia.org/wikipedia/commons/thumb/e/ea/Van_Gogh_-_Starry_Night_-_Google_Art_Project.jpg/1024px-Van_Gogh_-_Starry_Night_-_Google_Art_Project.jpg

import matplotlib.pyplot as plt
import matplotlib as mpl
import seaborn as sns
mpl.rcParams['figure.figsize'] = (10,10)
mpl.rcParams['axes.grid'] = False

import numpy as np
from PIL import Image
import time
import functools

from numpy import linalg as LA

%tensorflow_version 1.x
import tensorflow as tf

from tensorflow.python.keras.preprocessing import image as kp_image
from tensorflow.python.keras import models 
from tensorflow.python.keras import losses
from tensorflow.python.keras import layers
from tensorflow.python.keras import backend as K

TensorFlow 1.x selected.

tf.enable_eager_execution()
print("Eager execution: {}".format(tf.executing_eagerly()))

Eager execution: True

content_path = tf.keras.utils.get_file('nyc3.jpg', 'https://www.history.com/.image/ar_1:1%2Cc_fill%2Ccs_srgb%2Cfl_progressive%2Cq_auto:good%2Cw_1200/MTU3ODc5MDg1MzU3MDgyMzM1/new-york-city-3.jpg')
style_path = tf.keras.utils.get_file('cp2.jpg','https://i.pinimg.com/564x/0c/97/74/0c9774af605c455acb9dfdc2312e14a6.jpg')

def load_img(path_to_img):
  max_dim = 512
  img = Image.open(path_to_img)
  long = max(img.size)
  scale = max_dim/long
  img = img.resize((round(img.size[0]*scale), round(img.size[1]*scale)), Image.ANTIALIAS)
  
  img = kp_image.img_to_array(img)
  
  # We need to broadcast the image array such that it has a batch dimension 
  img = np.expand_dims(img, axis=0)
  return img

def imshow(img, title=None):
  # Remove the batch dimension
  out = np.squeeze(img, axis=0)
  # Normalize for display 
  out = out.astype('uint8')
  plt.imshow(out)
  if title is not None:
    plt.title(title)
  plt.imshow(out)

plt.figure(figsize=(10,10))

content = load_img(content_path).astype('uint8')
style = load_img(style_path).astype('uint8')

plt.subplot(1, 2, 1)
imshow(content, 'Content Image')

plt.subplot(1, 2, 2)
imshow(style, 'Style Image')
plt.show()

Image preprocessing

def load_and_process_img(path_to_img):
  img = load_img(path_to_img)
  img = tf.keras.applications.vgg19.preprocess_input(img)
  return img

def deprocess_img(processed_img):
  x = processed_img.copy()
  if len(x.shape) == 4:
    x = np.squeeze(x, 0)
  assert len(x.shape) == 3, ("Input to deprocess image must be an image of "
                             "dimension [1, height, width, channel] or [height, width, channel]")
  if len(x.shape) != 3:
    raise ValueError("Invalid input to deprocessing image")
  
  # perform the inverse of the preprocessing step
  x[:, :, 0] += 103.939
  x[:, :, 1] += 116.779
  x[:, :, 2] += 123.68
  x = x[:, :, ::-1]

  x = np.clip(x, 0, 255).astype('uint8')
  return x

extract features from selected layers

# Content layer where will pull our feature maps
content_layers = ['block5_conv2'] 

# Style layer we are interested in
style_layers = ['block1_conv1',
                'block2_conv1',
                'block3_conv1', 
                'block4_conv1', 
                'block5_conv1'
               ]

num_content_layers = len(content_layers)
num_style_layers = len(style_layers)

def get_model():
  """ Creates our model with access to intermediate layers. 
  
  This function will load the VGG19 model and access the intermediate layers. 
  These layers will then be used to create a new model that will take input image
  and return the outputs from these intermediate layers from the VGG model. 
  
  Returns:
    returns a keras model that takes image inputs and outputs the style and 
      content intermediate layers. 
  """
  # Load our model. We load pretrained VGG, trained on imagenet data
  vgg = tf.keras.applications.vgg19.VGG19(include_top=False, weights='imagenet')
  vgg.trainable = False
  # Get output layers corresponding to style and content layers 
  style_outputs = [vgg.get_layer(name).output for name in style_layers]
  content_outputs = [vgg.get_layer(name).output for name in content_layers]
  model_outputs = style_outputs + content_outputs
  # Build model 
  return models.Model(vgg.input, model_outputs)

compute loss

def get_content_loss(base_content, target):
  return tf.reduce_mean(tf.square(base_content - target))

def gram_matrix(input_tensor):
  # We make the image channels first 
  channels = int(input_tensor.shape[-1])
  a = tf.reshape(input_tensor, [-1, channels])
  n = tf.shape(a)[0]
  gram = tf.matmul(a, a, transpose_a=True)
  return gram / tf.cast(n, tf.float32)

def get_style_loss(base_style, gram_target):
  """Expects two images of dimension h, w, c"""
  # height, width, num filters of each layer
  # We scale the loss at a given layer by the size of the feature map and the number of filters
  height, width, channels = base_style.get_shape().as_list()
  gram_style = gram_matrix(base_style)
  
  return tf.reduce_mean(tf.square(gram_style - gram_target))# / (4. * (channels ** 2) * (width * height) ** 2)

def get_feature_representations(model, content_path, style_path):
  """Helper function to compute our content and style feature representations.

  This function will simply load and preprocess both the content and style 
  images from their path. Then it will feed them through the network to obtain
  the outputs of the intermediate layers. 
  
  Arguments:
    model: The model that we are using.
    content_path: The path to the content image.
    style_path: The path to the style image
    
  Returns:
    returns the style features and the content features. 
  """
  # Load our images in 
  content_image = load_and_process_img(content_path)
  style_image = load_and_process_img(style_path)
  
  # batch compute content and style features
  style_outputs = model(style_image)
  content_outputs = model(content_image)
  
  
  # Get the style and content feature representations from our model  
  style_features = [style_layer[0] for style_layer in style_outputs[:num_style_layers]]
  content_features = [content_layer[0] for content_layer in content_outputs[num_style_layers:]]
  return style_features, content_features

def compute_loss(model, loss_weights, init_image, gram_style_features, content_features):
  """This function will compute the loss total loss.
  
  Arguments:
    model: The model that will give us access to the intermediate layers
    loss_weights: The weights of each contribution of each loss function. 
      (style weight, content weight, and total variation weight)
    init_image: Our initial base image. This image is what we are updating with 
      our optimization process. We apply the gradients wrt the loss we are 
      calculating to this image.
    gram_style_features: Precomputed gram matrices corresponding to the 
      defined style layers of interest.
    content_features: Precomputed outputs from defined content layers of 
      interest.
      
  Returns:
    returns the total loss, style loss, content loss, and total variational loss
  """
  style_weight, content_weight = loss_weights
  
  # Feed our init image through our model. This will give us the content and 
  # style representations at our desired layers. Since we're using eager
  # our model is callable just like any other function!
  model_outputs = model(init_image)
  
  style_output_features = model_outputs[:num_style_layers]
  content_output_features = model_outputs[num_style_layers:]
  
  style_score = 0
  content_score = 0

  # Accumulate style losses from all layers
  # Here, we equally weight each contribution of each loss layer
  weight_per_style_layer = 1.0 / float(num_style_layers)
  for target_style, comb_style in zip(gram_style_features, style_output_features):
    style_score += weight_per_style_layer * get_style_loss(comb_style[0], target_style)
    
  # Accumulate content losses from all layers 
  weight_per_content_layer = 1.0 / float(num_content_layers)
  for target_content, comb_content in zip(content_features, content_output_features):
    content_score += weight_per_content_layer* get_content_loss(comb_content[0], target_content)

  # print('style score:',style_score, 'content score:',content_score)  
  style_score *= style_weight
  content_score *= content_weight

  # Get total loss
  loss = style_score + content_score 
  return loss, style_score, content_score

def compute_grads(cfg):
  with tf.GradientTape() as tape: 
    all_loss = compute_loss(**cfg)
  # Compute gradients wrt input image
  total_loss = all_loss[0]
  return tape.gradient(total_loss, cfg['init_image']), all_loss

output

import IPython.display

def run_style_transfer(content_path, 
                       style_path,
                       num_iterations=1000,
                       content_weight=1e3, 
                       style_weight=1e-2): # 1e-2
  # We don't need to (or want to) train any layers of our model, so we set their
  # trainable to false. 
  model = get_model() 
  for layer in model.layers:
    layer.trainable = False
  
  # Get the style and content feature representations (from our specified intermediate layers) 
  style_features, content_features = get_feature_representations(model, content_path, style_path)
  gram_style_features = [gram_matrix(style_feature) for style_feature in style_features]
  
  # Set initial image
  init_image = load_and_process_img(content_path)
  init_image = tf.Variable(init_image, dtype=tf.float32)
  # Create our optimizer
  opt = tf.train.AdamOptimizer(learning_rate=5, beta1=0.99, epsilon=1e-1)

  # For displaying intermediate images 
  iter_count = 1
  
  # Store our best result
  best_loss, best_img = float('inf'), None
  
  # Create a nice config 
  loss_weights = (style_weight, content_weight)
  cfg = {
      'model': model,
      'loss_weights': loss_weights,
      'init_image': init_image,
      'gram_style_features': gram_style_features,
      'content_features': content_features
  }
    
  # For displaying
  num_rows = 2
  num_cols = 5
  display_interval = num_iterations/(num_rows*num_cols)
  start_time = time.time()
  global_start = time.time()
  
  norm_means = np.array([103.939, 116.779, 123.68])
  min_vals = -norm_means
  max_vals = 255 - norm_means   
  
  imgs = []
  losses = []
  styles = []
  contents = []
  for i in range(num_iterations):
    grads, all_loss = compute_grads(cfg)
    loss, style_score, content_score = all_loss
    opt.apply_gradients([(grads, init_image)])
    clipped = tf.clip_by_value(init_image, min_vals, max_vals)
    init_image.assign(clipped)
    end_time = time.time() 
    
    if loss < best_loss:
      # Update best loss and best image from total loss. 
      best_loss = loss
      losses.append(best_loss)
      styles.append(style_score)
      contents.append(content_score)
      best_img = deprocess_img(init_image.numpy())

    if i % display_interval== 0:
      start_time = time.time()
      
      # Use the .numpy() method to get the concrete numpy array
      plot_img = init_image.numpy()
      plot_img = deprocess_img(plot_img)
      imgs.append(plot_img)
      IPython.display.clear_output(wait=True)
      IPython.display.display_png(Image.fromarray(plot_img))
      print('Iteration: {}'.format(i))        
      print('Total loss: {:.4e}, ' 
            'style loss: {:.4e}, '
            'content loss: {:.4e}, '
            'time: {:.4f}s'.format(loss, style_score, content_score, time.time() - start_time))
  print('Total time: {:.4f}s'.format(time.time() - global_start))
  IPython.display.clear_output(wait=True)
  plt.figure(figsize=(14,4))
  for i,img in enumerate(imgs):
      plt.subplot(num_rows,num_cols,i+1)
      plt.imshow(img)
      plt.xticks([])
      plt.yticks([])
      
  return best_img, best_loss, losses, contents, styles

result output

import random

random.seed(2021)

best, best_loss, losses, contents, styles = run_style_transfer(content_path, 
                                     style_path, num_iterations=1000)

f = plt.figure(figsize=(10,3))
ax = f.add_subplot(1, 1, 1)
ax.plot(losses, label = 'total losses')
ax.plot(contents, label = 'content losses')
ax.plot(styles, label = 'style losses')
ax.set_title('Loss with each iteration')
ax.legend()

<matplotlib.legend.Legend at 0x7fce5761f588>

Image.fromarray(best)

def show_results(best_img, content_path, style_path, show_large_final=True):
  plt.figure(figsize=(10, 5))
  content = load_img(content_path) 
  style = load_img(style_path)

  plt.subplot(1, 2, 1)
  imshow(content, 'Content Image')

  plt.subplot(1, 2, 2)
  imshow(style, 'Style Image')

  if show_large_final: 
    plt.figure(figsize=(10, 10))

    plt.imshow(best_img)
    plt.title('Output Image')
    plt.show()

show_results(best, content_path, style_path)

Replace MaxPooling with AveragePooling

# edit vgg19 from max_pooling to average_pooling
vgg = tf.keras.applications.vgg19.VGG19(include_top=False, weights='imagenet')
def replace_max_to_avg(model):
  block,number = 1,1
  renamed = tf.keras.Sequential()
  for layer in model.layers:
    if isinstance(layer, layers.MaxPooling2D):
      layer = layers.AveragePooling2D(layer.pool_size, layer.strides)
      block += 1
      number += 1
    renamed.add(layer)
  return renamed

vgg_avg = replace_max_to_avg(vgg)
vgg_avg.trainable = False
vgg_avg.summary()

Model: "sequential"
_________________________________________________________________
Layer (type)                 Output Shape              Param #   
=================================================================
block1_conv1 (Conv2D)        (None, None, None, 64)    1792      
_________________________________________________________________
block1_conv2 (Conv2D)        (None, None, None, 64)    36928     
_________________________________________________________________
average_pooling2d (AveragePo (None, None, None, 64)    0         
_________________________________________________________________
block2_conv1 (Conv2D)        (None, None, None, 128)   73856     
_________________________________________________________________
block2_conv2 (Conv2D)        (None, None, None, 128)   147584    
_________________________________________________________________
average_pooling2d_1 (Average (None, None, None, 128)   0         
_________________________________________________________________
block3_conv1 (Conv2D)        (None, None, None, 256)   295168    
_________________________________________________________________
block3_conv2 (Conv2D)        (None, None, None, 256)   590080    
_________________________________________________________________
block3_conv3 (Conv2D)        (None, None, None, 256)   590080    
_________________________________________________________________
block3_conv4 (Conv2D)        (None, None, None, 256)   590080    
_________________________________________________________________
average_pooling2d_2 (Average (None, None, None, 256)   0         
_________________________________________________________________
block4_conv1 (Conv2D)        (None, None, None, 512)   1180160   
_________________________________________________________________
block4_conv2 (Conv2D)        (None, None, None, 512)   2359808   
_________________________________________________________________
block4_conv3 (Conv2D)        (None, None, None, 512)   2359808   
_________________________________________________________________
block4_conv4 (Conv2D)        (None, None, None, 512)   2359808   
_________________________________________________________________
average_pooling2d_3 (Average (None, None, None, 512)   0         
_________________________________________________________________
block5_conv1 (Conv2D)        (None, None, None, 512)   2359808   
_________________________________________________________________
block5_conv2 (Conv2D)        (None, None, None, 512)   2359808   
_________________________________________________________________
block5_conv3 (Conv2D)        (None, None, None, 512)   2359808   
_________________________________________________________________
block5_conv4 (Conv2D)        (None, None, None, 512)   2359808   
_________________________________________________________________
average_pooling2d_4 (Average (None, None, None, 512)   0         
=================================================================
Total params: 20,024,384
Trainable params: 0
Non-trainable params: 20,024,384
_________________________________________________________________

def get_model():
  """ Creates our model with access to intermediate layers. 
  
  This function will load the VGG19 model and access the intermediate layers. 
  These layers will then be used to create a new model that will take input image
  and return the outputs from these intermediate layers from the VGG model. 
  
  Returns:
    returns a keras model that takes image inputs and outputs the style and 
      content intermediate layers. 
  """
  # Load our model. We load pretrained VGG, trained on imagenet data
  vgg = tf.keras.applications.vgg19.VGG19(include_top=False, weights='imagenet')
  vgg.trainable = False
  # Get output layers corresponding to style and content layers 
  style_outputs = [vgg_avg.get_layer(name).output for name in style_layers]
  content_outputs = [vgg_avg.get_layer(name).output for name in content_layers]
  model_outputs = style_outputs + content_outputs
  # Build model 
  return models.Model(vgg_avg.input, model_outputs)

best, best_loss, losses, styles, contents = run_style_transfer(content_path, 
                                     style_path, num_iterations=1000)

Image.fromarray(best)

Change content layers

content_layers = ['block4_conv2'] 

Custom weights for different style layers

def gram_matrix(input_tensor):
  # We make the image channels first 
  channels = int(input_tensor.shape[-1])
  a = tf.reshape(input_tensor, [-1, channels])
  n = tf.shape(a)[0]
  gram = tf.matmul(a, a, transpose_a=True)
  return gram / tf.cast(n, tf.float32)

def get_style_loss(base_style, gram_target):
  """Expects two images of dimension h, w, c"""
  # height, width, num filters of each layer
  # We scale the loss at a given layer by the size of the feature map and the number of filters
  height, width, channels = base_style.get_shape().as_list()
  gram_style = gram_matrix(base_style)
  
  return tf.reduce_mean(tf.square(gram_style - gram_target))# / (4. * (channels ** 2) * (width * height) ** 2)

def compute_loss(model, loss_weights, init_image, gram_style_features, content_features):
  """This function will compute the loss total loss.
  
  Arguments:
    model: The model that will give us access to the intermediate layers
    loss_weights: The weights of each contribution of each loss function. 
      (style weight, content weight, and total variation weight)
    init_image: Our initial base image. This image is what we are updating with 
      our optimization process. We apply the gradients wrt the loss we are 
      calculating to this image.
    gram_style_features: Precomputed gram matrices corresponding to the 
      defined style layers of interest.
    content_features: Precomputed outputs from defined content layers of 
      interest.
      
  Returns:
    returns the total loss, style loss, content loss, and total variational loss
  """
  style_weight, content_weight = loss_weights
  
  # Feed our init image through our model. This will give us the content and 
  # style representations at our desired layers. Since we're using eager
  # our model is callable just like any other function!
  model_outputs = model(init_image)
  
  style_output_features = model_outputs[:num_style_layers]
  content_output_features = model_outputs[num_style_layers:]
  
  style_score = 0
  content_score = 0

  # Accumulate style losses from all layers
  # Here, we equally weight each contribution of each loss layer
  # weight_per_style_layer = 1.0 / float(num_style_layers)
  weight_per_style_layer = tf.Variable([0.05, 0.2, 1, 0.2, 0.2],dtype = 'float32')
  i = 0
  for target_style, comb_style in zip(gram_style_features, style_output_features):
    style_score += weight_per_style_layer[i].numpy() * get_style_loss(comb_style[0], target_style)
    i += 1
    
  # Accumulate content losses from all layers 
  weight_per_content_layer = 1.0 / float(num_content_layers)
  for target_content, comb_content in zip(content_features, content_output_features):
    content_score += weight_per_content_layer* get_content_loss(comb_content[0], target_content)
  
  style_score *= style_weight
  content_score *= content_weight

  # Get total loss
  loss = style_score + content_score 
  return loss, style_score, content_score

import IPython.display

def run_style_transfer(content_path, 
                       style_path,
                       num_iterations=1000,
                       content_weight=1e3, 
                       style_weight=1e-3): # 1e-2
  # We don't need to (or want to) train any layers of our model, so we set their
  # trainable to false. 
  model = get_model() 
  for layer in model.layers:
    layer.trainable = False
  
  # Get the style and content feature representations (from our specified intermediate layers) 
  style_features, content_features = get_feature_representations(model, content_path, style_path)
  gram_style_features = [gram_matrix(style_feature) for style_feature in style_features]
  
  # Set initial image
  init_image = load_and_process_img(content_path)
  init_image = tf.Variable(init_image, dtype=tf.float32)
  # Create our optimizer
  opt = tf.train.AdamOptimizer(learning_rate=5, beta1=0.99, epsilon=1e-1)

  # For displaying intermediate images 
  iter_count = 1
  
  # Store our best result
  best_loss, best_img = float('inf'), None
  
  # Create a nice config 
  loss_weights = (style_weight, content_weight)
  cfg = {
      'model': model,
      'loss_weights': loss_weights,
      'init_image': init_image,
      'gram_style_features': gram_style_features,
      'content_features': content_features
  }
    
  # For displaying
  num_rows = 2
  num_cols = 5
  display_interval = num_iterations/(num_rows*num_cols)
  start_time = time.time()
  global_start = time.time()
  
  norm_means = np.array([103.939, 116.779, 123.68])
  min_vals = -norm_means
  max_vals = 255 - norm_means   
  
  imgs = []
  for i in range(num_iterations):
    grads, all_loss = compute_grads(cfg)
    loss, style_score, content_score = all_loss
    opt.apply_gradients([(grads, init_image)])
    clipped = tf.clip_by_value(init_image, min_vals, max_vals)
    init_image.assign(clipped)
    end_time = time.time() 
    
    if loss < best_loss:
      # Update best loss and best image from total loss. 
      best_loss = loss
      best_img = deprocess_img(init_image.numpy())

    if i % display_interval== 0:
      start_time = time.time()
      
      # Use the .numpy() method to get the concrete numpy array
      plot_img = init_image.numpy()
      plot_img = deprocess_img(plot_img)
      imgs.append(plot_img)
      IPython.display.clear_output(wait=True)
      IPython.display.display_png(Image.fromarray(plot_img))
      print('Iteration: {}'.format(i))        
      print('Total loss: {:.4e}, ' 
            'style loss: {:.4e}, '
            'content loss: {:.4e}, '
            'time: {:.4f}s'.format(loss, style_score, content_score, time.time() - start_time))
  print('Total time: {:.4f}s'.format(time.time() - global_start))
  IPython.display.clear_output(wait=True)
  plt.figure(figsize=(14,4))
  for i,img in enumerate(imgs):
      plt.subplot(num_rows,num_cols,i+1)
      plt.imshow(img)
      plt.xticks([])
      plt.yticks([])
      
  return best_img, best_loss 

best, best_loss = run_style_transfer(content_path, 
                                     style_path, num_iterations=1000)

Image.fromarray(best)

Adjusting $\alpha:\beta$

style_path = tf.keras.utils.get_file('s1.jpg', 'https://i.pinimg.com/736x/e0/43/41/e04341dc9fd7b34ff1b1c06d14df164b.jpg')
content_path = tf.keras.utils.get_file('c1.jpg','https://i.pinimg.com/564x/57/16/b7/5716b7b65c955183c4ad3b6ea94a660b.jpg')

import IPython.display

def run_style_transfer(content_path, 
                       style_path,
                       num_iterations=1000,
                       content_weight=1e3, # initial 1e3
                       style_weight=1e-2): # initial 1e-2
  # We don't need to (or want to) train any layers of our model, so we set their
  # trainable to false. 
  model = get_model() 
  for layer in model.layers:
    layer.trainable = False
  
  # Get the style and content feature representations (from our specified intermediate layers) 
  style_features, content_features = get_feature_representations(model, content_path, style_path)
  gram_style_features = [gram_matrix(style_feature) for style_feature in style_features]
  
  # Set initial image
  init_image = load_and_process_img(content_path)
  init_image = tf.Variable(init_image, dtype=tf.float32)
  # Create our optimizer
  opt = tf.train.AdamOptimizer(learning_rate=5, beta1=0.99, epsilon=1e-1)

  # For displaying intermediate images 
  iter_count = 1
  
  # Store our best result
  best_loss, best_img = float('inf'), None
  
  # Create a nice config 
  loss_weights = (style_weight, content_weight)
  cfg = {
      'model': model,
      'loss_weights': loss_weights,
      'init_image': init_image,
      'gram_style_features': gram_style_features,
      'content_features': content_features
  }
    
  # For displaying
  num_rows = 2
  num_cols = 5
  display_interval = num_iterations/(num_rows*num_cols)
  start_time = time.time()
  global_start = time.time()
  
  norm_means = np.array([103.939, 116.779, 123.68])
  min_vals = -norm_means
  max_vals = 255 - norm_means   
  
  imgs = []
  losses = []
  for i in range(num_iterations):
    grads, all_loss = compute_grads(cfg)
    loss, style_score, content_score = all_loss
    opt.apply_gradients([(grads, init_image)])
    clipped = tf.clip_by_value(init_image, min_vals, max_vals)
    init_image.assign(clipped)
    end_time = time.time() 
    
    if loss < best_loss:
      # Update best loss and best image from total loss. 
      best_loss = loss
      losses.append(best_loss)
      best_img = deprocess_img(init_image.numpy())

    if i % display_interval== 0:
      start_time = time.time()
      
      # Use the .numpy() method to get the concrete numpy array
      plot_img = init_image.numpy()
      plot_img = deprocess_img(plot_img)
      imgs.append(plot_img)
      IPython.display.clear_output(wait=True)
      IPython.display.display_png(Image.fromarray(plot_img))
      print('Iteration: {}'.format(i))        
      print('Total loss: {:.4e}, ' 
            'style loss: {:.4e}, '
            'content loss: {:.4e}, '
            'time: {:.4f}s'.format(loss, style_score, content_score, time.time() - start_time))
  print('Total time: {:.4f}s'.format(time.time() - global_start))
  IPython.display.clear_output(wait=True)
  plt.figure(figsize=(14,4))
  for i,img in enumerate(imgs):
      plt.subplot(num_rows,num_cols,i+1)
      plt.imshow(img)
      plt.xticks([])
      plt.yticks([])
      
  return best_img, best_loss 

best, best_loss = run_style_transfer(content_path, 
                                     style_path, num_iterations=1000,
                                     content_weight=1e-2, style_weight=1e3)

# with alpha/beta = 1e-5
Image.fromarray(best)

Final Output

With content_layer = block4_2, lr = 5, iteration = 200, weights_per_style_layer = (0.15,10,0.2,10,0.2),

content_path = tf.keras.utils.get_file('rainy.jpg', 'https://i.pinimg.com/564x/57/16/b7/5716b7b65c955183c4ad3b6ea94a660b.jpg')
style_path = tf.keras.utils.get_file('cp1.jpg','https://i.pinimg.com/474x/7d/4e/e5/7d4ee518d817e409f6df35ba212a19b2.jpg')
content_path2 = tf.keras.utils.get_file('nyc3.jpg', 'https://www.history.com/.image/ar_1:1%2Cc_fill%2Ccs_srgb%2Cfl_progressive%2Cq_auto:good%2Cw_1200/MTU3ODc5MDg1MzU3MDgyMzM1/new-york-city-3.jpg')
style_path2 = tf.keras.utils.get_file('cp2.jpg','https://i.pinimg.com/564x/0c/97/74/0c9774af605c455acb9dfdc2312e14a6.jpg')

Downloading data from https://i.pinimg.com/564x/57/16/b7/5716b7b65c955183c4ad3b6ea94a660b.jpg
40960/40890 [==============================] - 0s 0us/step
Downloading data from https://i.pinimg.com/474x/7d/4e/e5/7d4ee518d817e409f6df35ba212a19b2.jpg
122880/116756 [===============================] - 0s 0us/step

content_layers = ['block4_conv2'] 
start_time = time.time()
best, best_loss = run_style_transfer(content_path, 
                                     style_path, num_iterations=200,
                                     content_weight = 1000, style_weight = 1)
print('Time:', time.time()-start_time)
print('Total Loss:', best_loss)

Time: 23.382643461227417
Total Loss: tf.Tensor(1643762800.0, shape=(), dtype=float32)

Image.fromarray(best)

# Compare with the base model with equal weights on style images
content_layers = ['block5_conv2'] 
start_time = time.time()
best, best_loss = run_style_transfer(content_path, 
                                     style_path, num_iterations=1000,
                                     content_weight = 1000, style_weight = 1e-2)
print('Time:', time.time()-start_time)
print('Total Loss:', best_loss)

Time: 111.73320722579956
Total Loss: tf.Tensor(1780934.4, shape=(), dtype=float32)

Image.fromarray(best)

content_layers = ['block4_conv2'] 
start_time = time.time()
best, best_loss= run_style_transfer(content_path, 
                                     style_path2, num_iterations=300,
                                     content_weight = 1e3, style_weight = 1)
print('Time:', time.time()-start_time)

Time: 34.4533531665802

Image.fromarray(best)

content_layers = ['block4_conv2'] 
start_time = time.time()
best, best_loss = run_style_transfer(content_path2, 
                                     style_path, num_iterations=200,
                                     content_weight = 1e3, style_weight = 1)
print('Time:', time.time()-start_time)

Time: 36.25394058227539

Image.fromarray(best)

content_layers = ['block4_conv2'] 
start_time = time.time()
best, best_loss= run_style_transfer(content_path2, 
                                     style_path2, num_iterations=200,
                                     content_weight = 1e3, style_weight = 1)
print('Time:', time.time()-start_time)

Time: 35.67541027069092

Image.fromarray(best)